- Rating
- 2
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper proposes Variable Elimination Networks for recovering graph states from sparse observations. Shared neural networks parameterize local Gaussian potentials, which are assembled into a positive-definite precision matrix. Observations enter through Gaussian likelihood factors, and differentiable elimination produces conditional means and covariances. The paper establishes basic structural properties and evaluates the method across simulated and measured recovery tasks, with additional analyses of extrapolation, uncertainty, and repeated-query efficiency.
Critical strengths
-
The architecture provides a coherent formulation of graph-wide conditional recovery, with guaranteed positive definiteness and interchangeable inference backends.
-
The evaluation is extensive, including five-seed results, physical estimators, component ablations, uncertainty analysis, and detailed latency measurements.
-
The paper reports its limitations transparently, including weaker results on native DGMRF tasks and restrictions associated with caching, graph structure, and distribution shift.
Critical weaknesses
-
The methodological advance over existing learned Gaussian models needs sharper differentiation. Exact differentiable inference with learned Gaussian potentials was already explored by [1] in their work on deep Gaussian CRFs. The theoretical propositions here mainly establish standard properties of the construction.
-
The observation interface is ambiguous. Appendix B states that the source network receives all node channels, apparently including observations that are subsequently inserted as likelihood factors. This complicates the prior/posterior interpretation and the meaning of the “No observation” ablation.
-
The DGMRF comparison changes several components simultaneously. DGMRF uses a linear mean and Gaussian-likelihood training, whereas VEN uses a nonlinear source network and primarily MSE training. Consequently, the results do not isolate the contribution of the proposed precision parameterization.
[1] Chandra, Siddhartha, and Iasonas Kokkinos. "Fast, exact and multi-scale inference for semantic image segmentation with deep gaussian crfs." European conference on computer vision. Cham: Springer International Publishing, 2016.
Questions for authors
-
Does the source network receive the observed state values used in the likelihood? Are these values also removed in the “No observation” ablation?
-
Are there existing comparisons that match the source-network capacity and training objective between VEN and DGMRF? How much of the reported advantage remains under those controls?
-
Which precision and observation-noise parameters are learned under MSE, and what determines the absolute covariance scale? Appendix E partially addresses calibration, but this aspect of the probabilistic interpretation remains unclear.
2 comments
Thank you for clarifying which inputs enter the source network. The task-specific explanation resolves my concern about the same target measurement being used twice and makes the “No observation” ablation interpretable. I still view the contribution as an architectural advance over existing Gaussian CRFs, and I would limit the conclusion from the matched comparison to GasLib.
Clarifying VEN's Observation Interface, Controls and Architectural Contribution
Thank you. We clarify the observation interface, controlled comparison, covariance scale and architectural contribution.
GasLib uses boundary-flow measurements/indicators and static attributes in the source; pressure observations enter through the likelihood. SimBench/CINELDI use P/Q pseudo-measurements, their masks and static attributes; voltage/angle PMUs enter separately. LeakDB/AQI use time/calendar context for the prior mean; EEGBCI learns a prior mean. Thus the prior is conditional on available context, before incorporating target-state evidence.
"No observation" removes the target-state likelihood, returning μ0=Q−1b (or m when b=Qm). Target observations are already absent from the source; context covariates remain. It therefore measures the contribution of target-state sensors given those covariates. Five passing interface checks on archived models and inspection of data construction/forward paths support this separation. We will clarify the ablation's definition.
Matched source capacity and training objective. We added a GasLib comparison using the authors' DGMRF/FlexLayer precision operator with the same one-hidden-layer SiLU source-MLP architecture, unobserved-node MSE, observation likelihood, data/splits/masks, optimizer, training budget and validation-selection rule. The frozen five-seed test gives NRMSE 0.04828±0.00072 for VEN versus 0.54057±0.15527 for DGMRF under this shared protocol; VEN is lower in all five pairs. Stored best-validation checkpoints were evaluated in float64 without retraining or reselection. Across ten checkpoints, precision- and covariance-form posterior means agreed within 2.1×10−10 in the numerical audit. The gap persists after controlling the source-capacity and objective differences raised in the review.
Learned parameters and covariance scale. MSE backpropagates through the conditional mean into the active precision-factor networks, source/prior parameters and observation precision. GasLib/EEGBCI/AQI learn scalar node/edge terms; LeakDB learns edge/boundary terms; SimBench/CINELDI learn 2×2 edge blocks with reference-bus anchoring. Observation precision is τ=ϵ+softplus(ρ): one shared scalar for scalar tasks, one per state channel for SimBench/CINELDI. Positivity floors and reference constraints are fixed.
With Q≻0 and R≻0, these parameters define a normalized conditional Gaussian. Calibration concerns agreement with the data distribution. Mean MSE alone does not identify absolute covariance scale. Writing Q+=Q+S⊤R−1S, we have μ=Q+−1(b+S⊤R−1y) and Σ=Q+−1. Whenever common rescaling is permitted, (Q,b,R)↦(cQ,cb,R/c) for c>0 leaves μ unchanged and maps Σ to Σ/c. Initialization, constraints and regularization can select a numerical scale, but do not establish calibration.
Appendix E already reports uncertainty evaluation. For unobserved validation entries V, the scale is α=∣V∣−1∑i∈V(zi−μi)2/Σii, with Σcal=αΣ and unchanged mean (channelwise for SimBench). In the reported marginal-NLL training setting, SimBench raw NLL changes from 0.139 to −0.520 and 95% coverage from 99.96% to 96.54%, with NRMSE 0.17855→0.18017 (five-seed means; different training budgets). We will explicitly distinguish exact inference within the learned Gaussian model from empirical uncertainty calibration.
Chandra-Kokkinos [1] predicts A(I),B(I) from each image and solves [A(I)+λId]s=B(I) for pixel-class scores. VEN specifies different factor generators, parameter sharing and input dependencies for graph-state recovery. Relative to local MPNNs, elimination permits conditioning beyond a fixed hop radius; relative to [1], the distinction is the learned factor architecture and conditional-query computation.
[1] Chandra & Kokkinos. Fast, Exact and Multi-Scale Inference for Semantic Image Segmentation with Deep Gaussian CRFs. ECCV 2016.